Herbal medicines: what the evidence says

A best-selling immune remedy.

Tribulus is one of the most popular testosterone boosters and one of the least supported: controlled trials in healthy men consistently show no increase in testosterone, strength or lean mass.

Rhodiola is one of the better-studied adaptogens, but that is a low bar: the fatigue trials are small and mostly at high risk of bias, and the one rigorous depression trial found it no better than placebo.

Sold as an adaptogen for stamina and stress resistance, but the best controlled trial found no overall benefit for chronic fatigue.

Popular for libido and "hormone balance", but every systematic review reaches the same verdict: a handful of small trials, largely from the same research groups, too weak to conclude anything.

Small trials consistently report benefits for stress and blood glucose, but that consistency is itself a red flag: literally every published human study is favourable, most are small, short and industry-linked, and none are large or independently replicated.

An adaptogen with a large Soviet-era literature and almost no modern controlled evidence.

The "brain tonic" claims do not survive meta-analysis, which found no benefit in any cognitive domain.

The largest and longest trial ever run on ginkgo found it did nothing: over a median 6.

Small trials suggest lemon balm acutely increases self-rated calmness and may reduce anxiety and depression scores, but the pooled evidence rests on a few tiny, highly heterogeneous studies.

Genuinely interesting pharmacology - kanna's alkaloids inhibit serotonin reuptake and PDE4 - but the entire human evidence base is three tiny studies totalling under 80 people.

Single-dose studies show small improvements in speed of thinking, but the one 12-week trial found nothing at all.

Valerian is the most-studied herbal sleep aid and still has not clearly beaten placebo on objective sleep measures.

Passiflora incarnata has a handful of small trials suggesting mild calming and modest subjective sleep benefit, but almost nothing at the scale or quality needed to call it effective.

Hops appears in nearly every herbal sleep formula, almost always alongside valerian, and the standalone human evidence is a couple of tiny pilot studies.

Magnolia bark's active compounds honokiol and magnolol have genuinely interesting GABA-A activity in animals, but the human evidence is a couple of tiny industry-funded pilot studies on a Magnolia-Phellodendron blend.

The most-studied cold remedy in the supplement aisle, and still an unresolved one: the Cochrane review of 24 trials found no reliable treatment effect and only a small, statistically fragile prevention signal.

Almost the entire clinical literature on Astragalus comes from Chinese trials of injectable decoctions added to conventional drugs, not from the oral capsules sold in the West.

Olive leaf extract shows small, inconsistent improvements in blood pressure and lipids across a thin set of mostly short trials — nowhere near the antiviral and immune claims made for it in marketing.

Oregano oil is a genuinely potent antimicrobial in a petri dish and almost untested in humans — one 14-person open-label parasite study is the most-cited clinical evidence there is.

Goldenseal has no credible clinical evidence for any of the infections it is sold for, and it is one of the strongest herbal CYP3A4 inhibitors ever measured in humans — meaning its best-documented biological effect is a drug interaction, not a benefit.

Two small randomized trials from the early 2000s found modest pain relief in knee osteoarthritis and rheumatoid arthritis, and essentially nothing has confirmed them in the two decades since.

A genuinely nutrient-dense leaf whose supplement claims don't hold up: a 2025 GRADE-assessed meta-analysis found no consistent cardiometabolic benefit and rated the certainty of evidence as very low for every single outcome.

Bergamot polyphenols show promising cholesterol and glucose reductions, but almost all the positive data come from a small cluster of trials by overlapping research groups in one region.

Olive polyphenols demonstrably protect LDL from oxidation — the one claim with regulatory backing.

Cinnamon produces small reductions in fasting glucose in type 2 diabetes, but the effect on HbA1c — the outcome that actually matters — is inconsistent across meta-analyses.

Gymnema's most reliable effect is a curious one: its gymnemic acids temporarily switch off your ability to taste sweetness.

A Cochrane review found no significant glucose-lowering benefit for bitter melon versus placebo, and none versus metformin or glibenclamide.

White kidney bean extract really does inhibit starch-digesting amylase, but the weight effect in humans is small and the trial base is weak: an early meta-analysis found no significant weight difference versus placebo, and a more recent one found about 1.

Green coffee extract became a bestseller on the back of a trial that was later retracted and a marketing campaign that drew regulatory enforcement.

The best-designed trial of Garcinia cambogia — a 135-person randomised study published in JAMA — found no weight or fat loss beyond placebo.

Guarana works, but almost certainly because it is a caffeine delivery system with 2-4 times the caffeine of coffee beans by weight.

A caffeinated polyphenol-rich tea with modest, inconsistent effects on blood lipids and no convincing weight-loss data.

Anthocyanin-rich blueberry preparations show small, inconsistent cognitive benefits, mostly in older adults who already have some decline.

The coffee cherry around the bean is polyphenol-rich and has a small set of mostly manufacturer-funded trials showing modest, inconsistent cognitive effects.

Deglycyrrhizinated licorice removes the compound that raises blood pressure, which makes it far safer than whole licorice — but the efficacy evidence is mostly small trials from the 1970s-80s, largely superseded by acid suppression, plus a handful of modern trials of a different, flavonoid-standardised extract.

Despite decades of use as a "liver protector," milk thistle has repeatedly failed its best-designed trials: a Cochrane review found no mortality or histology benefit in alcoholic and viral liver disease, and two well-run NASH trials missed their primary endpoints.

Triphala, the classic Ayurvedic blend of amalaki, bibhitaki and haritaki, has its best (though still low-quality) evidence as a mouthwash for plaque and gingivitis, where it performs roughly like chlorhexidine.

Fennel has repeatedly reduced menstrual pain in small randomized trials — meta-analyses find it beats placebo and performs roughly like mefenamic acid — but almost every study is small, short and single-region, so certainty stays low.

Devil's claw shows a real but small and low-confidence signal for short-term low back pain at 50-100 mg harpagoside per day.

Polypodium leucotomos extract raises the UV dose needed to redden skin and may reduce photodermatosis flares, but trials are small and short and the protection is weak — equivalent to a very low SPF.

Feverfew is a long-popular migraine preventive whose best trial produced a difference of only 0.

Nettle root produced a statistically significant but clinically trivial improvement in prostate symptom scores in its largest trial, with no change in urine flow or residual volume.

Horsetail has a genuine, measurable diuretic effect — one small crossover trial found 900 mg/day matched hydrochlorothiazide 25 mg — but that is essentially the entire human evidence base.

Despite decades of use for hot flushes, the pooled randomized evidence does not show black cohosh beats placebo.

Red clover isoflavones produce at best a small reduction in hot flushes — around one to two flushes per day more than placebo in pooled trials — which reviewers have questioned as falling short of clinical meaningfulness.

Shatavari (Asparagus racemosus) is a staple Ayurvedic women's tonic with a very thin modern trial record — a couple of small randomized studies in muscle strength and lactation, and essentially nothing rigorous for menopause or fertility.

Wild yam is sold as a "natural progesterone" alternative, and that premise is simply wrong: the human body cannot convert its diosgenin into progesterone or DHEA.

Uva ursi (bearberry) has been tested head-to-head as an antibiotic-sparing treatment for uncomplicated UTI and lost — worse symptom resolution and more cases of pyelonephritis in the herbal arm.

Saw palmetto is the best-studied prostate supplement and the trials are consistently negative: 32 randomized trials in over 5,600 men show it performs no better than placebo for urinary symptoms, even at double and triple the standard dose.

The one large, year-long, placebo-controlled trial of pumpkin seed for BPH missed its primary endpoint.

Horny goat weed (Epimedium) has no credible human trials for erectile dysfunction — the entire premise is in-vitro PDE5 inhibition by icariin that is orders of magnitude weaker than sildenafil.

Isoflavone-standardized kudzu extract has repeatedly reduced how much alcohol heavy drinkers consume in small trials - but nearly all of them come from one research group.

A standardized celery seed extract lowered blood pressure substantially in a small crossover trial, but essentially all the human data come from one research team.

Almost all human data on Cissus come from a small cluster of trials testing proprietary multi-ingredient products, so the eye-catching weight-loss numbers cannot be credited to Cissus itself.

Forskolin's fat-burner reputation rests on two small 12-week trials from 2005 - and neither actually produced weight loss.

Bitter orange (Citrus aurantium) and its alkaloid p-synephrine became the post-ephedra fat burner ingredient, but a systematic review and meta-analysis found essentially no meaningful weight-loss effect while confirming measurable cardiovascular activity.

Hoodia was the appetite suppressant of the 2000s, backed by a compelling San Bushmen story and almost no clinical data.

Dong quai (Angelica sinensis) is marketed as a "female ginseng" for hot flashes and menstrual complaints, but the one well-designed placebo-controlled trial of dong quai alone found it no better than placebo.

Rehmannia glutinosa is a foundational "kidney yin" herb in Chinese medicine and is almost never studied alone — the human evidence is for multi-herb formulas like Rehmannia-6 (Liu Wei Di Huang Wan), where results are modest and the herb's individual contribution is unknown.

He shou wu (Polygonum/Reynoutria multiflora) is sold for hair regrowth and longevity on the strength of a folk legend, not clinical trials.

Codonopsis pilosula (dang shen) is used in Chinese medicine as a gentler, cheaper substitute for ginseng, and virtually all of its human evidence comes from multi-herb formulas rather than the herb alone.

Panax notoginseng (san qi, tienchi) is a distinct species from Asian ginseng with a different saponin profile and a genuinely different use: it is given as an injectable stroke drug in China (Xuesaitong) with a large but low-quality trial base.

Cistanche ("desert ginseng") is heavily marketed for testosterone and libido, but the actual human trials tested mobility in older adults and cognition in combination with ginkgo — and found modest effects.

Eucommia ulmoides bark (du zhong) has one small clinical trial showing a modest blood pressure reduction, and essentially nothing else in humans.

Goji berries are a genuinely nutrient-dense food with an unusually high zeaxanthin content, and small trials show they raise macular pigment density.

Guduchi (giloy) is an Ayurvedic "immune booster" whose human evidence is thin and whose main claim to fame in the modern literature is a wave of drug-induced autoimmune-like hepatitis cases reported during the COVID-19 pandemic.

Arjuna bark is the classic Ayurvedic heart herb, and it is well tolerated, but the best-designed trial to date failed its primary endpoint in heart failure.

Neem has real antimicrobial activity and the best human data are for neem mouthrinse in gingivitis, where small trials put it roughly on par with chlorhexidine.

Guggul is the textbook case of an herb that looked good in early Indian trials and then failed in the West: a JAMA-published RCT found guggulipid raised LDL cholesterol rather than lowering it.

Ashitaba is a Japanese herb whose yellow sap chalcones (4-hydroxyderricin, xanthoangelol) have one small positive RCT for visceral fat.

Banaba leaf standardized to corosolic acid produces short-term reductions in blood glucose in small human studies, but the trials are tiny, mostly industry-run and decades old, with no HbA1c-level evidence.

Sarsaparilla is a root sold as a "blood purifier" and testosterone booster with essentially zero controlled human evidence for either.

Wild lettuce ("lettuce opium") is marketed as a natural opioid alternative, which it is not — its lactones are sedative in mice at high doses and have no affinity for opioid receptors.

Boldo is a Chilean leaf used as a digestive and "liver" tea with no controlled human efficacy trials and documented cases of hepatotoxicity — the opposite of what it is sold to do.

Gentian root is the archetypal bitter, one of the most intensely bitter substances used in food, and it does reliably stimulate salivary and gastric secretion.

Bitters tinctures — gentian, wormwood, dandelion, artichoke, angelica in alcohol — act on a real mechanism: bitter taste receptors in the mouth and gut modulate digestive secretion and gut hormones.

A Brazilian bark (Erythrina mulungu/verna) sold as a natural anxiolytic and sleep aid, but the only two human trials are in dental surgery patients and the larger, better-designed one found it no better than placebo.

Brazil's best-known gastric herb (Monteverdia/Maytenus ilicifolia) has one modern randomized trial showing it is non-inferior to omeprazole for reflux dyspepsia — but that trial had no placebo arm, and commercial product adulteration is well documented.

A bitter Brazilian herb (Baccharis trimera) sold for liver detox, blood sugar, and weight loss with zero human randomized trials behind any of those claims.

Marketed as 'vegetable insulin,' Bauhinia forficata has one clearly negative trial of the traditional tea and one positive trial of a standardized 300 mg extract.

Brazil's most famous aphrodisiac bark has no controlled trial of catuaba alone for sexual function — the only human data come from Catuama, a four-herb combination, and they concern burning mouth syndrome and basic safety.

The 'potency wood' of the Amazon has decent rodent memory and antidepressant data and a clean 28-day tolerability record as part of a combination product, but not a single controlled human trial of muira puama alone for libido, erections, or cognition.

A Mexican shrub sold for female libido on the strength of tradition alone — no controlled trial has ever tested damiana for sexual function.

Sold as a 'Russian secret' anabolic because it contains ecdysteroids, but a typical 500 mg capsule delivers only a few milligrams of them — far below the doses used in the one human ecdysterone trial.

Heavily marketed as a natural cancer treatment on the basis of cell-culture data, graviola has never been tested for cancer in a human trial — and its signature compounds are the leading suspect in an epidemic of atypical parkinsonism in Guadeloupe.

Noni juice is among the most aggressively marketed cure-all supplements and among the least supported: after decades of sales there is still no adequately powered trial showing benefit for any condition, and there is a documented signal for idiosyncratic liver injury.

Mangosteen pericarp xanthones are pharmacologically interesting and mangosteen gel has decent data as a local periodontal treatment — but the juice sold through multi-level marketing for "immune and inflammation support" has essentially no oral-supplement evidence behind it.

The famous "RAF pilots ate bilberry jam to see in the dark" story is wartime myth, and controlled trials have repeatedly failed to find a night-vision benefit.

Green tea extract is the single most common ingredient in "fat burner" formulas, and the Cochrane review found it produces essentially no weight loss — a pooled difference of 0.

Marshmallow root's mucilage forms a soothing film over an irritated throat, which is a plausible mechanism but has never been tested against placebo in a proper randomized trial.

Mullein is a traditional respiratory herb that has essentially never been tested as an oral supplement in a controlled human trial.

Rosemary is an excellent food antioxidant with a famous but thin cognitive literature: the one controlled dose-ranging trial found 750 mg of rosemary powder helped memory speed slightly while 6000 mg made it worse.

Standardized spearmint extract (the rosmarinic-acid-rich Neumentix ingredient) has produced small, statistically significant attention improvements in a handful of trials — all funded by the ingredient's manufacturer.

White horehound is a traditional cough candy and bitter tonic whose two testable claims have both failed when checked: the FDA removed it from the approved OTC expectorant list for lack of evidence, and the one randomized diabetes trial found it lowered blood glucose by 0.

Elecampane root has no randomized human trials for anything.

Clove oil is a legitimate topical dental analgesic thanks to its eugenol content, and that is where its evidence begins and ends.

Myrrh has a defensible niche as a topical antiseptic mouth rinse and essentially nothing else: the one time a myrrh drug was tested head-to-head as an antiparasitic, it failed badly.

Oregon grape cream has modest randomized evidence for mild-to-moderate plaque psoriasis, where it beats placebo but loses to standard prescription topicals.

Chinese skullcap (Scutellaria baicalensis) has one week-long trial for knee osteoarthritis using a proprietary blend, and a documented history of liver injury when its baicalin was sold as the medical food flavocoxid.

A mild traditional sedative whose only decent human trial tested it inside a three-ingredient combination, where it beat placebo by under 2 points on a 56-point anxiety scale.

The sleep evidence belongs to jujube seed (suan zao ren) inside multi-herb Chinese formulas, not to the sweet jujube fruit most supplements sell.

A bitter, inulin-rich root sold for liver 'detox' and water retention with essentially no supporting human trials — the single clinical diuretic study used the leaf, not the root, in 17 people with no control group.

One small Iranian trial found burdock root tea lowered inflammatory and oxidative markers in knee osteoarthritis, but it never reported pain or function.

The 'stone breaker' does not break stones.

This is where almost all commercial resveratrol comes from, so its longevity claims inherit resveratrol's unimpressive human record.

Artemisinin derivatives are among the great drugs of modern medicine — for malaria, given as prescription medicines.

A polyphenol-rich Mediterranean rock rose with one placebo-controlled cold trial from 2009 that has never been replicated, plus cell-culture antiviral data that gets over-extrapolated into systemic immune claims.

A Siberian ecdysterone source with a Soviet sports-science legend behind it and almost no modern human data.

Pycnogenol is a patented French maritime pine bark extract with dozens of small trials spread thinly across a dozen unrelated conditions.

Nopal (Opuntia ficus-indica) is a genuinely nutritious food whose soluble fiber modestly binds dietary fat and blunts glucose, but supplement-dose extracts produce small and inconsistent effects.

A traditional North American demulcent bark whose mucilage plausibly coats and soothes irritated mucosa, but which has essentially never been tested on its own in a controlled human trial.

A resin from the Chios mastic tree with one genuinely decent RCT in functional dyspepsia and a weak, much-overhyped signal against H.

Corydalis yanhusuo is a genuinely interesting analgesic herb: its alkaloids act on dopamine receptors and produce real pain relief in animal models without opioid tolerance.

A traditional European herb (Euphrasia) named for its folk use in eye complaints.

A traditional "partus preparator" tea taken in late pregnancy, but the only decent randomised trial found no shortening of labour's first stage and only a ~10-minute effect on the second.

A bitter herb sold mainly as a breastfeeding "galactagogue," almost always blended with fenugreek — and it has never been tested alone in a controlled human trial.

A West African shrub that became a testosterone-booster trend on the strength of a single 2005 rat study — with zero human trials of any kind.

An old "blood purifier" and green-food tablet whose only respectable human data — an LDL reduction — came from eating around 120 g of alfalfa seeds a day, not from tablets.

A traditional North American nervine herb (Scutellaria lateriflora) used for anxiety and restlessness.

A pleasant traditional bedtime tea made from Tilia flowers, long used in Europe and Latin America for tension and restlessness.

Famous for its effect on cats, catnip (Nepeta cataria) is also sold as a mild human sedative tea.

A bitter Artemisia herb with a long history as a digestive tonic and, in folklore, as a dream herb.

An Ayurvedic root (Nardostachys jatamansi) related to valerian and used for sleep and stress.

An Egyptian water lily (Nymphaea caerulea) sold as a relaxing tea or vape, with claims built almost entirely on ancient iconography.

A European nervine (Stachys officinalis) traditionally used for tension headaches and nervous unease.

Known in Chinese medicine as the 'collective happiness bark' and used for low mood and disturbed sleep.

A pleasant Chinese tea flower traditionally drunk for tired, red or dry eyes.

Roasted Cassia obtusifolia seeds, brewed as a tea in Chinese medicine for eye strain and constipation.

An orchid used in Chinese medicine as a yin tonic for dryness, including dry eyes.

A root used as the chief herb in classical Chinese formulas for stress and low mood.

Sold almost exclusively paired with magnolia bark in a proprietary extract marketed for cortisol and stress.

A traditional European herb for palpitations and nervous tension.

A Chinese root traditionally used for memory and restlessness.

An Ayurvedic brain tonic, usually Convolvulus pluricaulis.

The 'intellect tree' of Ayurveda, sold as a seed oil for memory.

An Amazonian holly leaf brewed as a caffeinated tea.

A West African seed chewed for alertness and once used to flavour cola drinks.

A Mediterranean herbal tea studied for cognition and neuroprotection.

South American bark containing lapachol and beta-lapachone.

A Mediterranean mint-family herb with a long tradition in cough and cold remedies.

A Rocky Mountain root used in Native American and herbalist traditions for coughs and sore throats.

A North American herb once widely used for fevers and influenza.

A core cooling herb in Chinese medicine for sore throat and fever.

One of the most-used herbs in Chinese medicine for colds and inflammation, and one of the most-studied in the laboratory.

A hugely popular over-the-counter cold remedy in China, sold mainly as granules.

A pungent herb eaten as a vegetable and used medicinally across East Asia, with a strong laboratory record for antiviral and anti-inflammatory activity.

A traditional Ayurvedic herbal jam built around amla, taken daily for general vitality and immunity.

A mint-family herb used in Chinese medicine for thyroid nodules and, in the West, for herpes.

A root used by Indigenous peoples of the American West and marketed as a natural antiviral.

Glossy privet fruit (nu zhen zi), a Chinese tonic usually paired with astragalus.

The spice is the industrial source of shikimic acid, the starting material for oseltamivir, but taking star anise does not produce an antiviral drug in your body.

Horseradish root, usually combined with nasturtium, is a licensed herbal remedy in Germany for airway and urinary infections.

A rosmarinic-acid-rich leaf extract studied for seasonal allergies.

A South African herb (cancer bush) promoted for immunity, HIV and fatigue.

The Ayurvedic fruit best known as a component of Triphala, taken for regularity and detox.

An Ayurvedic mix of black pepper, long pepper and ginger used as a digestive and as a bioenhancer.

Pippali is an Ayurvedic spice rich in piperine.

A bitter herb whose only reasonable human trial is in Crohn's disease, not parasites.

A traditional parasite-cleanse ingredient, usually paired with wormwood and clove.

Chinese rhubarb contains anthraquinones, the same stimulant laxative class found in senna.

A Chinese herb whose active compound is berberine.

Bai zhu is one of the most-used herbs in Chinese digestive formulas but is almost never tested on its own.

A common weed used as a soothing herb for coughs and the gut lining.

Chen pi is a standard Chinese digestive herb rich in citrus flavonoids.

A culinary spice with a handful of small metabolic trials, mostly from Iran.

A carminative spice used for bloating and cramping.

A traditional Chinese root used for angina and circulation.

A traditional Chinese root used for headache and blood flow.

The fruit of Panax ginseng, richer in ginsenoside Re than the root.

The culinary bay leaf tested as a blood sugar supplement.

Seed extract of Irvingia gabonensis, marketed on the back of trials reporting dramatic weight loss.

An Indian edible cactus traditionally used to blunt hunger during long journeys.

A Brazilian herb (Cordia ecalyculata) sold for appetite suppression and water weight.

A West African spice (Aframomum melegueta) that activates brown fat and modestly raises energy expenditure in small human studies.

A common garden weed with a long folk reputation as a slimming and diuretic herb.

Blends of dandelion, uva ursi, juniper and similar herbs sold to shed water weight.

An Ayurvedic root (Rubia cordifolia) traditionally used as a blood purifier for skin complaints.

An Ayurvedic herb famous for hair growth and liver support.

The plant that gave aspirin its name, containing salicylate precursors.

Marigold flower extract with a modest evidence base for topical wound and skin care.

A ginger relative whose joint reputation comes from one large trial where it was combined with ginger.

A triterpene extract whose only well-tested use is a topical gel for wound healing in the rare skin disease epidermolysis bullosa.

A traditional East Asian and North American rhizome used for tendons and joints.

A saponin-rich desert root long sold for arthritis.

An Amazonian tree resin rich in beta-caryophyllene, marketed for inflammation and skin.

A concentrated cucumber extract sold for joint pain.

A caffeine-free South African tea rich in aspalathin.

A traditional bleeding herb with two small Iranian randomized trials behind it.

A traditional antispasmodic bark with real smooth-muscle activity in the laboratory but essentially no controlled human data for menstrual cramps.

A uterine antispasmodic whose entire human record consists of 19th-century clinical reports.

A tannin-rich herb long used for heavy periods, but its only human studies are topical ones for mouth ulcers and wounds.

A classical Ayurvedic uterine tonic with centuries of use and almost no controlled human data.

An Ayurvedic bark used for menstrual complaints and PCOS.

A Malaysian women's tonic with a handful of small trials, none of which establishes benefit for menopausal symptoms.

A culinary herb tested once, as a tea, in women with PCOS.

An Ayurvedic root sold for libido and vitality.

A traditional Chinese seed whose coumarin osthole is well characterized in the laboratory and untested in people for erectile function.

A yang tonic sold for libido and stamina, with no human trials for either.

A Thai root with unusually potent phytoestrogens.

A traditional galactagogue with no trial of its own.

A Thai tuber root traditionally used for male sexual function.

A South African herb sold as a testosterone booster on the strength of rat studies.

A traditional diuretic and urinary herb.

A culinary herb with a long history as a diuretic.

The staple of heavy-metal detox protocols.

A traditional lymphatic and urinary herb.

A European herb used to flush the urinary tract, approved for that traditional use by German regulators.

The silky styles of maize, used traditionally as a soothing diuretic for bladder irritation.

A South African leaf long used for urinary complaints.

A traditional herb for fevers, wounds and bleeding.

A bitter root traditionally used as a gentle laxative and a tonic for iron.

A mint-family herb used in Germany for mild overactive thyroid symptoms.

An Ayurvedic herb (Boerhavia diffusa) used traditionally as a diuretic and kidney tonic.

One of the three fruits in triphala, promoted for cholesterol and respiratory health.

A Himalayan bitter root used in Ayurveda for the liver.

A traditional Chinese "yin tonic" root used for dry cough and dry throat.

Bedtime blends of valerian, chamomile, lemon balm, passionflower and lavender.

A classical Chinese multi-herb formula (also sold as Gui Pi Wan) for insomnia with worry and fatigue.

The classic Chinese insomnia decoction, built around sour jujube seed.

A traditional Chinese formula for restless sleep with palpitations and anxiety.

Whole bark from the African tree Pausinystalia yohimbe, sold as a natural alternative to pharmaceutical yohimbine.

An aromatic African root (white ginger) used traditionally as an aphrodisiac and appetite stimulant.

A parasitic desert plant used in Chinese medicine as a yang tonic for stamina and sexual function.

A North American root used traditionally to induce labour and ease menstrual cramps.

A slow-growing, now-threatened American woodland root used in Western herbal medicine for irregular cycles and fertility.

A classic eight-herb Chinese formula combining the qi tonic Si Jun Zi Tang with the blood tonic Si Wu Tang, used for fatigue and post-partum or post-illness depletion.

The four-herb blood tonic that underlies most Chinese gynaecological formulas, used for irregular or painful periods.

Slimming teas built around senna and other stimulant laxatives, sold as detoxification.

The leaf of Nelumbo nucifera, used in Chinese medicine and sold as a slimming tea for its quercetin glycosides.

Semi-fermented tea whose catechins and caffeine raise fat oxidation slightly in small metabolic studies.

Microbially fermented tea popular in China for weight and cholesterol.

A traditional Chinese rhizome used for headache and dizziness, whose main compound gastrodin has been tested for migraine.

A Chinese herb used traditionally for tremor, convulsions and hypertension, and best known in the West as an ingredient of the Kampo formula yokukansan.

A traditional herb whose alkaloid lobeline resembles nicotine and was long promoted for quitting smoking.

An aromatic culinary herb tested mostly as an aroma rather than a supplement.

A classical Chinese formula prescribed for stress-related complaints.

Multi-herb antimicrobial formulas, usually berberine plus oregano or thyme oil.

An Ayurvedic fruit traditionally used for diarrhoea and dysentery.

Bark of Holarrhena antidysenterica, used in Ayurveda for dysentery and intestinal parasites.

A carminative Indian spice used for bloating and indigestion.

A pungent resin used as a digestive spice.

A sweet carminative seed.

An intensely bitter Himalayan herb used in Ayurveda for digestion and fever.

Seed of the sensitive plant, sold in gut-cleanse and parasite protocols because its mucilage forms a sticky gel.

Nutgrass rhizome, used in Ayurveda and Chinese medicine for bloating and diarrhoea.

A classical Ayurvedic multi-herb powder for acidity and reflux.

A traditional fennel, ginger and bicarbonate infant preparation for colic.

The flower of Sambucus nigra, used traditionally as a sweat-inducing remedy for fevers and colds.

Eriobotrya japonica leaf, a fixture of Chinese and Japanese cough syrups such as pipa gao.

An Ayurvedic herb for cough and bronchitis whose alkaloid vasicine inspired the mucolytic drugs bromhexine and ambroxol.

A classical Chinese formula built on honeysuckle and forsythia, taken at the first sign of a cold.

A three-herb Chinese formula (astragalus, atractylodes, siler) used to reduce the frequency of infections.

A classical Ayurvedic powder of rock sugar, bamboo silica, long pepper, cardamom and cinnamon for cough and congestion.

Multi-herb capsules and tinctures marketed for viral season, usually combining echinacea, elderberry, andrographis, olive leaf and garlic.

The least processed form of Camellia sinensis, marketed as the most antioxidant-rich tea.

Multi-herb protocols built on Japanese knotweed, cat's claw, cryptolepis and andrographis, sold for persistent Lyme symptoms.

A classical Ayurvedic multi-herb tablet (Bauhinia variegata bark plus guggul resin) sold for thyroid nodules and lymph swelling.

A six-herb Chinese formula used for centuries as a kidney-yin tonic.

Solomon's seal rhizome, a Chinese tonic used for fatigue and dryness, rich in polysaccharides and steroidal saponins.

Garlic slowly heat-aged until it turns dark and sweet, sold for cholesterol and antioxidant effects.

Ficus carica leaf, traditionally taken as a decoction for diabetes in Mediterranean folk medicine.

Guava leaf tea, approved in Japan as a food for specified health use for post-meal blood sugar.

Syzygium cumini fruit and seed powder, one of the best known Ayurvedic remedies for diabetes.

A classical Ayurvedic decoction of ten roots, used traditionally for pain, inflammation and vata disorders.

A fern rhizome used in Chinese medicine for fractures and bone loss.

A root used in Chinese formulas for the knees and lower back, almost always as one component of a multi-herb prescription.

Standardised for 6-MSITC, an isothiocyanate found in real wasabi rhizome.

Guaco syrup is one of Brazil's most popular cough remedies and is even supplied by the public health system, yet not a single controlled trial has tested whether it works.